The long-term objective of this proposed research is to develop retroviral vectors which deliver therapeutic genes for diabetes specifically to liver and pancreatic beta-cells. The short-term aim of this project is to alter the cell-binding site of the retroviral surface Envelope protein (Env) such that it only mediates binding and entry of the retrovirus specifically into liver or beta-cells. This research develops the use of the feline leukemia virus (FeLV) Env as the backbone for modification. The FeLV Env is more amenable for this modification than previously studied Env proteins, such as those of the marine leukemia viruses, because the FeLV Env encodes a short stretch of amino acids within its amino terminus that determines receptor specificity. One targeting strategy employs a library of 106-107 env genes with random amino acids substituted into this receptor-determining region. The library will be screened for the ability of am of the random sequences to enable entry specifically into liver or beta-cells. Preliminary results indicate that Env proteins with novel targeting specificities can be derived from such a selection screen. A second strategy is to use a library with a liver-specific peptide substituted into the receptor-determining region. In this approach, the liver-targeting peptide is flanked by random amino acids in order to optimize the conformation of the peptide. This library will be screened for the specific targeting of liver cells. These strategies differ from earlier approaches to alter retroviral targeting by the use of bulky antibody fragments or other large cell-binding ligands. The short peptide sequences being introduced into the FeLV Env in the project described here should only minimally perturb the Env protein. These specifically targeted Env proteins will eventually be used to deliver genes that can ether protect beta cells from immune destruction or induce insulin expression in liver cells.